Impacts of ecological restoration projects on agricultural productivity in China
Post on 13-Dec-2016
Embed Size (px)
J. Geogr. Sci. 2013, 23(3): 404-416 DOI: 10.1007/s11442-013-1018-6
2013 Science Press Springer-Verlag
Received: 2012-10-19 Accepted: 2012-11-06 Foundation: National Key Program for Developing Basic Science, No.2010CB950904; National Natural Science Founda-
tion of China, No.41071344; Knowledge Innovation Program of CAS, No.KZCX2-EW-306; Strategic Prior-ity Research Program of CAS, No.XDA05050602
Author: Qin Yuanwei (1983), Ph.D, specialized in the study of land use/cover change and its eco-environmental effects. E-mail: email@example.com
*Corresponding author: Yan Huimin, Ph.D, E-mail: firstname.lastname@example.org
Impacts of ecological restoration projects on agri-cultural productivity in China
QIN Yuanwei1,2, *YAN Huimin1, LIU Jiyuan1, DONG Jinwei3, CHEN Jingqing1, XIAO Xiangming3 1. Institute of Geographic Sciences and Natural Resources Research, CAS, Beijing 100101, China; 2. University of Chinese Academy of Sciences, Beijing 100049, China; 3. Departments of Microbiology and Plant Botany, Center for Spatial Analysis, University of Oklahoma, Norman,
Oklahoma 73019-5300, USA
Abstract: The changes in cropland quantity and quality due to land use are critical concerns to national food security, particularly for China. Despite the significant ecological effects, the ecological restoration program (ERP), started from 1999, has evidently altered the spatial patterns of Chinas cropland and agricultural productivity. Based on cropland dynamic data from 2000 to 2008 primarily derived from satellite images with a 30-m resolution and satel-lite-based net primary productivity models, we identified the impacts on agricultural produc-tivity caused by ERP, including Grain for Green Program (GFGP) and Reclaimed Cropland to Lake (RCTL) Program. Our results indicated that the agricultural productivity lost with a rate of 132.67104 t/a due to ERP, which accounted for 44.01% of the total loss rate caused by land use changes during 20002005. During 20052008, the loss rate due to ERP de-creased to 77.18104 t/a, which was equivalent to 58.17% of that in the first five years and 30.22% of the total loss rate caused by land use changes. The agricultural productivity loss from 20002008 caused by ERP was more attributed to GFGP (about 70%) than RCTL. Al-though ERP had a certain influence on cropland productivity during 20002008, its effect was still much less than that of urbanization; moreover, ERP was already converted from the project implementation phase to the consolidation phase.
Keywords: ecological restoration; agricultural productivity; remote sensing; Grain for Green; Reclaimed Crop-land to Lake
In history, land use/cover change was dominated by substantial increase of cropland and built-up land and great decrease of forest to meet increasing resources requirement of human
QIN Yuanwei et al.: Impacts of ecological restoration projects on agricultural productivity in China 405
being to some extent, which impacted the agricultural productivity by changing the quantity, quality and land use structure of cropland resources (Foley et al., 2005). In China, food se-curity has always been a concern because of the challenge of lack of cropland, increasing population and water shortage (Tao et al., 2009). During the past 50 years, remarkable achievement in agricultural production was reached, although China is facing a great chal-lenge of land scarcity to feed the largest population with cropland per capita far below the world average (Chen, 2007). The cereal production has increased steadily with an annual growth rate of 3.7%, which is substantially higher than the world mean growth rate of 2% during the period (Fan et al., 2012). Although the great promotion in cereal production mainly resulted from the yield increase, it was still attributed to cropland expansion, espe-cially in Northeast and Northwest China (Liu et al., 2005, 2009; Wang et al., 2009). Ac-cording to National Bureau of Statistics (NBS, hereafter), Chinas cropland area increased by about 30% during 19782000 (http://www.stats.gov.cn/tjsj/ndsj/). Based on the cropland dynamics monitoring through Landsat TM/ETM images at a spatial resolution of 30 m, cropland increased about 2.79 million ha in China during 19902000, which was mainly in the Northeast and Northwest regions and was primarily due to reclamation of grassland and deforestation (Liu et al., 2005). However, under tremendous pressure on land and food de-mand, excessive cropland reclamation had resulted in a series of ecological and environ-mental problems that offset a large part of the acquired achievement (Shi et al., 2011). Most of the primary forest and wetland in China has been depleted, and a high percentage of new cultivated land and grassland has been degraded (WWF, 2003; Yin et al., 2005). Unreason-able cropland reclamation exacerbated water shortage in the north area (State Council of Peoples Republic of China, 2008) and the newly added cropland always had poor quality (Liu et al., 2005; Dong et al., 2010). Excessive wetland reclamation shrunk water area, in-duced soil degradation and deteriorated the stability of regional ecosystem significantly (Li et al., 2006; Zheng et al., 2006). Compared with cropland, afforested area had enhanced vegetation structure, species diversity, soil nutrients and anti-erodibility (Jiao et al., 2010; Li et al., 2010), and increased storages of soil organic carbon and nitrogen (Liu et al., 2004), just like the grassland restoration (Wang et al., 2011). National level ecological restoration program (ERP) was triggered in China by severe droughts in 1997 and huge floods in 1998.
ERP program include Grain for Green Program (GFGP) (Zhang et al., 1999; Loucks et al., 2001; Xu et al., 2006; Liu et al., 2008) and Reclaimed Cropland to Lake (RCTL, hereafter) Programs. ERP is one of the worlds largest ecological restoration programs and plays an important role in global conservation efforts. After pilot in Sichuan, Shaanxi and Gansu in 1999, ERP was widely carried out in 2000. During the first 5 years, the ERP was dominated by ecological construction that a large area of cropland not suitable for cultiva-tion was reversed to ecological land, such as forest, grassland and wetland. Due to the great change in supply-demand relationship and increasing food price in the international grain market in 2003, the attention of ERP was gradually turned to the consolidation of the recov-ered ecological land after 2005 (Huang et al., 2010; State Council of Peoples Republic of China, 2007), and the cropland loss rate caused by ERP slowed down. Cropland database at the scale of 1:100,000, derived from Landsat images with a 30-m resolution, could clearly depict the spatial and temporal patterns and dynamics of Chinas cropland since the end of the 1980s (Liu et al., 2003, 2009), particularly ERP. Every coin has two sides; the conver-
406 Journal of Geographical Sciences
sion of cropland to ecological land potentially affected the agricultural production. Although the effects of ecological restoration on agricultural production could be not severe, as its negative effects were offset by new land reclamation in Northeast and Northwest China (Deng et al., 2005, 2006), more concern is needed to evaluate the impacts or implications of these ecological restoration programs on food security of China. Remote sensing is increas-ingly used in monitoring agricultural productivity and land use dynamics (including defor-estation and afforestation, cropland reclamation and abandonment, urban expansion, etc.) (Doraiswamy et al., 2003; Tao et al., 2005; Liu et al., 2005; Morton et al., 2006; Yan et al., 2009; Gibbs et al., 2010), which enable large-scale and real-time monitoring cropland area and agricultural productivity. Agricultural productivity of different crops were expressed as accumulated dry matter in net primary productivity (NPP), which could provide a unified measure standard for crop productivity; so, it is an effective and feasible measuring index for agricultural productivity change analysis. Satellite-based light use efficiency models have been an important and widely accepted method to calculate ecosystem NPP (Potter et al., 1993; Prince et al., 1995; Lobell et al., 2002). Therefore, the regional agricultural pro-duction could be estimated by cropland area and NPP.
This study aimed to estimate the ERPs impacts on agricultural productivity during 20002005 and 20052008 on national scale, by combining satellite-based light use effi-ciency models with cropland dynamics due to ERP. The impacts of GFGP and RCTL on ag-ricultural productivity were distinguished to well understand the spatial and temporal pat-terns and regional discrepancies of the reduced agricultural productivity in China.
2 Data and methods 2.1 Cropland change data
In order to investigate the impact of land use change process on cropland resources across China, a research team led by the author Liu J Y, has carried out national Land-Use/Cover Change monitoring through remote sensing since the early 1990s. The state and change of cropland was identified as the core of the monitoring. The National Land-Use/Cover Change Data sets (NLCD, hereafter) were also developed based on satellite images and a variety of other data including soil type, DEM, roads, rivers and climate. The state and dynamic grid data contained information on the percent area of cropland with a resolution of 1 km1 km was obtained since the end of the 1980s (Liu et al., 2002, 2005, 2009). Each grid land-use vector data were acquired first by remote sensing images interpretation through a com-puter-aided, interactive procedure (Liu et al., 2003, 2005). Then, the vector data and a vector fishnet with 1 km 1 km cells were intersected, and the area percentage of land-use dynam-ics in every cell was calculated and finally, the cells were converted into raster grid, con-tained information of area percentage of land-use dynamics. This aggregation process can help for not only effective data fusion but also maintaining the acreage information without information loss (Liu et al., 2005). The cropland data in 2000 was primarily interpreted from Landsat TM/ETM images in 1999/2000, while cropland data in 2005 and 2008 were inter-preted from Landsat TM/ETM images and CBERS images in 2004/2005 and 2007/2008, respectively.
QIN Yuanwei et al.: Impacts of ecological restoration projects on agricultural productivity in China 407
To characterize the ERPs impact on cropland productivity, we used data converted from cropland in this study, including cropland converted to forest, cropland to grassland and cropland to water body during 20002005 and 20052008. In the NLCD, cropland is defined as identifiable reclaimed cropland in remote sensing images; while in the investigation rules of Ministry of Land and Resources, new reclamation of wasteland to be cultivation for more than 3 years could be identified as cropland.
2.2 Agricultural productivity data
GLO-PEM is a productivity efficiency model driven mainly by National Oceanic and At-mospheric Administration (NOAA) Advanced Very High Resolution Radiometers (AVHRR) data. The model consists of several interrelated components about the processes of canopy radiation absorption, utilization, autotrophic respiration and the regulation of these processes by environmental factors. The structure principle of GLO-PEM as well as its application in agricultural productivity estimation in China was discussed in detail (Yan et al., 2009). VPM (Vegetation Photosynthesis Model) is an ecosystem productivity estimation model based on MODIS (Moderate Resolution Imaging Spectroradiometer) data (Yan et al., 2007, 2012). The climate data that drive VPM model came from daily surface climate dataset during 20002005 provided by China Meteorological Data Sharing Service System of National Meteorological Administration (http://cdc.cma.gov.cn/). The average temperature index of daily data of 752 ground meteorological stations and automatic stations were used in this study through spline interpolation by using ANUspline software.
2.3 Impacts of ERP on agricultural productivity
The agricultural production per unit area estimated from the satellite-based ecosystem productivity models and cropland dataset with a spatial resolution of 1 km in grid cell pro-vided precise distribution and area information. So it is possible to calculate the total agri-cultural productivity based on the raster data containing cropland area information and the estimated production per unit area which was represented as the average net primary pro-ductivity (NPP) during 19822005. The total agricultural productivity of two periods was calculated according to equation (1), respectively.
ANPPcP = (1) where P is the total agricultural productivity (Ton C); NPP is the cropland production per unit area (gC/m2/yr); c is the proportion of cropland in each grid cell, and A is the grid area. The change of total agricultural productivity (P) is equal to the changed productivity caused by cropland area change (A) and cropland production per unit area.
To investigate the difference of ecological restoration effects on agricultural productivity in spatial and temporal patterns across China during 20002008, we divided the land of China into 8 regions (Figure 1): Northeast China Plain region ( 1) including: Heilongjiang, Jilin and Liaoning; Huang-Huai-Hai Plain region ( 2) including: Beijing, Tianjin, Hebei, Shandong and Henan; Middle-lower Yangtze Plain region ( 3) including: Shanghai, J i-angsu, Anhui, Hubei, Hunan, Zhejiang and Jiangxi; South region ( 4) including: Guan g-dong and Fujian; Northern arid and semi-arid region ( 5) including: Xinjiang , Inner Mon-golia, Ningxia and Gansu; Loess Plateau region ( 1) including: Sha anxi and Shanxi; Si-chuan Basin and surrounding regions ( 2) including: Sichuan and Chongqing;
408 Journal of Geographical Sciences
Figure 1 Chinas agricultural regionalization. Northeast China Plain region includes: 11 -Heilongjiang,
12 -Jilin and 13 -Liaoning; Huang-Huai-Hai Plain region includes: 21 -Beijing, 22 -Tianjin, 23 -Hebei, 24 -Shandong and 25 -Henan; Middle-lower Yangtze Plain region includes: 31 -Shanghai, 32 -Jiangsu, 33 -Anhui, 34 -Hubei, 35 -Hunan, 36 -Jiangxi and 37 -Zhejiang; South region includes: 41 -Guangdong
and 42 -Fujian; Northern arid and semi-arid region includes: 51 -Xinjiang, 52 -Inner Mongolia, 53 -Gansu and 54 -Ningxia; Losses Plateau region includes: 11 -Shaanxi and 12 -Shanxi; Sichuan Basin and surrounding regions includes: 21 -Sichuan and 22 -Chongqing; Yunnan-Guizhou Plateau includes: 31 -Yunnan,
32 -Guizhou and 33 -Guangxi.
Yunnan-Guizhou Plateau region ( 3) including: Yunnan, Guizhou and Guangxi. 1 - 5 were the major agricultural production regions and 1 - 3 were the major regions under the implementation of G...